Why does Firefox treat Helvetica differently from Chrome? was marked as “not constructive” without explanation. I can only speculate that it was perceived as impossible to answer with any kind of factual evidence. If so, that perception is incorrect:

  1. W3C standards are public. I would like to know what standards (if any) are relevant to this question. Are the browsers ignoring a clear standard (and therefore buggy)? If we need a new standard where would it go? If there’s an existing standard that needs clarification, what is it?

  2. Firefox and Chrome are open source projects. It is not a matter of speculation as to how they render text. If none of us have the tenacity to dig around in the source, please don’t blame our collective inexperience on the question.

  • 6
    Stack Overflow isn't a discussion forum, and questions about the inner workings of browsers are beyond the site scope. If you can rephrase your question so that it addresses a specific programming problem, i'll reopen.
    – user102937
    Feb 1, 2013 at 18:37
  • Regarding point 2 - that could just as easily be used as a negative. After all, they're both open source, so there's no speculation, or, indeed, question as to why they may behave differently. I didn't see the original question, or vote, but I suspect that may be a contributory factor.
    – RivieraKid
    Feb 1, 2013 at 18:46
  • 1
    Looks like it's been reopened by the community. Feb 1, 2013 at 19:15
  • So it has! Feel free to mark this as status-completed. Much as I appreciate @GeorgeStocker taking the time to explain his view, I disagree. Hopefully he doesn’t take it the wrong way if I don’t accept it.
    – Adam
    Feb 1, 2013 at 19:27
  • 2
    Oh I'm a little bit hesitant on doing that, in case the question ends up being closed again. I will, however, post an answer of my own here. (And maybe, just maybe, I'll research the topic on my own anyway.) Feb 1, 2013 at 19:32
  • 1
    @Adam Doesn't matter to me whether or not you accept it. :-) There are times that the answer the community likes the most isn't the accepted answer. The accepted answer has no bearing on what the 'official' stance is (though on meta, I sometimes wish it did). Feb 1, 2013 at 20:15

3 Answers 3


As you've stated, your question is quite confusing:

Why is there a discrepancy? Is it because of a lack of standards for web typography, or is this a bug that can be filed against one or the other browser?

Also, I apologize for making this a multi-part question, which seems to have caused some confusion. More important than how to work around this discrepancy is why it exists.

This question is about why the height of the content area of a text block and the vertical position of the text within it varies between browsers. Specifically, what standards regulate the relationship between the content area and the baselines and descender lines of a font? How is the line-height attribute supposed to affect the height of the content area? Even if there is no such standard, how does type setting work in these browsers?

I can tell what you're asking, but only amidst a pretty messy description (it's hard to pick out specific parts of the multi-part question that you describe it to be). You may need to rewrite these paragraphs into something more concise, concrete, and to the point.

And while I'm here I'll toss in my two cents: I don't think this is really "not constructive". In fact, I find this question as useful and reasonable for Stack Overflow as questions are on topic as web development things, despite them not being true programming languages.

Sure, it probably won't give rise to any directly actionable answers, but what it does offer in answers is understanding of why something works this way or doesn't work that way. That understanding, in my books, can be useful for practical purposes. Besides, the answer to this is likely to be objective as it ought to derive from something explicitly stated in the specifications, which you're already looking for in an answer anyway.

In any case, I'm going to research the topic you've asked about and see what turns up. Even if it's decided that the question isn't a good fit for the site, I'm as interested to know the answer as you are.

  • Agreed. It’s messy. Would the question be more in line with the guidelines if it were broken into two separate questions? For example: #1 Are there any standards that regulate X behavior? #2 How do Chrome and Firefox define the size of content areas and the position of the text within them?
    – Adam
    Feb 1, 2013 at 21:12
  • @Adam: Hmmm... I think it's fine to keep them together as it is (Chrome and Firefox behaving differently was what prompted your question to begin with). It's just the bits I quoted that are messy. I have a feeling that #2 would be just as poorly-received if not even more anyway, because it's based on implementation details that aren't often very useful to authors. (Which, now that I mention it, is probably what the others are referring to.) Feb 1, 2013 at 21:15
  • 3
    @Adam, I've made some edits that I think streamline this considerably. Also, thanks for the analysis, BoltClock - I think you're dead-on.
    – Shog9
    Feb 1, 2013 at 22:08

This question is a great question and it's well researched.

But... In its current state, it doesn't belong on Stack Overflow.

Here's why:

  • It is asking why where no reasonable person can actually answer it. Unless the Chrome developers or Mozilla developers see fit to answer it, it's essentially unanwerable. I can provide my opinions as to why, but that's all it is -- an opinion. That's what makes the question not Constructive.

  • Let's say the answer is "Lack of Standards" -- Does that really help you solve whatever problem you face? If the answer is no, then that's a sign that the question itself is not constructive. More importantly, Does it matter to the problem you face whether it's a lack of standards or a bug? It doesn't. What really matters is fixing the issue. That's what questions should be focused on, not navel-gazing as to why the world isn't perfect.

Stack Overflow is meant to be a place for actionable answers to actual problems people face.

If you had asked 'How do I work around this issue", that would be a good fit for Stack Overflow.

If you update the question to narrow the scope to how to fix a specific problem you face, then I'll be happy to re-open it.

EDIT: Shog9 made some good edits to the question, and re-opened it. In the form he edited it to, I think it's a good fit for Stack Overflow (knowing that the actual content has not changed, but rather how it was presented, and how the question was focused to something that was answerable).

  • 1
    Good analysis. :)
    – Alenanno
    Feb 1, 2013 at 18:36
  • 8
    How do you define reasonable? Is looking at the source of a tool really unreasonable in answering a question about how to use it? If “lack of standards” really is the answer, it helps in two ways: 1) Other people with the same question at least have an answer instead of a mysteriously closed question—even if that problem can’t be helped. 2) It helps a broader problem, which is how do we improve the tools with which we work? Our voice is stronger as a community, especially when that voice says, “This is broken.”
    – Adam
    Feb 1, 2013 at 19:03
  • 6
    @Adam 1) is circular reasoning. "If the question got answered then we'd have an answer." Of course we would. As for 2), this is not what Stack Overflow is here for. We're here to answer specific programming questions, not rally the community around a cause. (Although I don't really think that was the point of your original question to begin with.) Feb 1, 2013 at 19:17
  • 5
    @BilltheLizard: 1) Having an unpleasant answer is more helpful than having a closed question. There’s nothing circular about that. 2) If a question is perceived by a few people as unanswerable, shouldn’t we let it remain unanswered instead of burying it?
    – Adam
    Feb 1, 2013 at 19:34
  • 7
    -1 This is exactly the kind of mentality that is slowly making SO an inhospitable place for actually interesting questions and reducing it to a crowd-sourced debugging service. Feb 1, 2013 at 21:52
  • @NullUserExceptionอ_อ I invite you to post an answer to this Meta question then: meta.stackexchange.com/questions/165952/… Feb 2, 2013 at 2:44
  • 1
    -1 for "What really matters is fixing the issue." That's not always the case.
    – TRiG
    Feb 14, 2013 at 11:25

You said a dirty word: w*y.

A large part of the Stack Overflow community hates it when you ask why something happens. This phenomenon is specific to Stack Overflow, I haven't seen it on other Stack Exchange sites where I participate (such as Unix & Linux or Security, to stick to technical fields). Those sites are ticking just fine, producing valuable content retaining experts, having markedly less strife around than on SO. So it's definitely U&L and Sec.SE doing it right, and SO doing it wrong.

So why (look, SOpedians! I said “why”!) is it so bad to ask why? Here's my $0.02 sociological analysis. (Warning: if you paid that price, you've been swindled.)

To ask why is disruptive. “Don't make me think” is well-known advice in user interface design, and why is all about making one think. To ask not only how to do this but why is this so demonstrates curiosity, a search for more, a willingness to outreach yourself.

Yet such propensity is common among artists, scientists and craftsmen. Aren't programmers craftsmen?

Some are, some aren't. I think among the ones who are, a second factor is at play: the myth of the hacker. A hacker does not write a program: a hacker exudes a program. If you need to code by a spec, you aren't a real hacker. To ask why is to second-guess yourself, because it puts the design into question — but the design must flow.

So the yin and the yang of the programmer conspires to reject the idea that there may be intelligence in the design. It's all in the execution, baby!

Why was it done that way? Who knows? Who cares? I don't want to care! Don't make me think!

Of course, if you start thinking about what you're doing, if you value understanding over doing by rote, that does make you a better programmer. But that's neither here nor there. Stack Overflow is not really about good programming questions and answers. It's about popular questions and answers. Help me get a fish plzsendtehcodez kthxbye. Start asking why, and that's a question that requires more than two br*in¹ cells to answer. That takes thinking, and knowledge! No avenue for quick rep! No, that cannot stand on SO.

Once upon a time, there was no “non constructive” close reason. There was, instead, “subjective and argumentative”. The close reason was renamed because people tended to be overzealous and close any subjective question. “Not constructive” was supposed to make it explicit that the problem was questions that did not call for “answers to be supported by facts, references, or specific expertise”. The gist of it is that a bad question invites answers that say “I like the red one”, whereas a good question invites answers that teach you something — why the proposed solution is a good one.

This hasn't worked: a question that asks why is rather likely to call on “specific expertise”, yet “non-constructive” is used as a “this is subjective” close reason. Not that why is particularly subjective — it's only subjective if you don't know why — but why does look subjective if you don't know why, and for any given questions, most people don't know why.

It is, sadly, difficult to get a why question accepted on Stack Overflow. Calling on specific expertise is frowned upon. My subjective impression is that this problem is getting worse over time: the SO community (or at least, the part that takes interest in its governance) is predominantly interested in volume over quality, and SO has such a volume of questions that it cannot fail to be successful by that measure. The flip side of the coin is that most questions are crap. Of course, 90% of everything is crap. Yet it saddens me that SO is rejecting some of its best elements. I hope it won't take all of SE down with it.

¹ Also a dirty word on SO, which explains why mentioning Br*infuck by name is risky on MSO.

  • 6
    I could write an entire answer just refuting the allegations you've made here. I think you're seeing one thing, and taking it as another. As someone who pays attention to how Stack Overflow is governed, I can say that I've never thought any of those things you've claimed in this post. Feb 1, 2013 at 21:28
  • 1
    Questions of the form, "Why did X do Y." Tend to not be well received, and rightly so, because only X can answer that question. Questions of the form, "Why should I do Y" can be answered by anyone, and tend to be much more well received on SO.
    – Servy
    Feb 1, 2013 at 21:28
  • @Gilles Well, that's the reason given for it's closure so if you feel it's not appropriate, then by all means address that. Whether or not that's what the question is really asking, that's what the readers have perceived it to be asking, and they have reacted accordingly. Perhaps the question can be improved to avoid that misconception.
    – Servy
    Feb 1, 2013 at 21:35
  • 1
    @Servy: Hence my answer. Feb 1, 2013 at 21:36
  • @Servy I don't feel that the question is unsuitable for SO in any way in its present state. So I am limiting my action on it to voting to reopen. Feb 1, 2013 at 21:42
  • @Gilles And that's fine. I'm addressing your meta post; I feel that your answer to the meta question is deeply flawed. The reasoning for your actions is incorrect, not the proposed action itself.
    – Servy
    Feb 1, 2013 at 21:45
  • @Gilles I invite you to post an answer to this meta question (especially since I called out your post by name): meta.stackexchange.com/questions/165952/… Feb 2, 2013 at 2:44
  • I quite agree with your third section here, but I don't share your ideas about the myth of the hacker. To me at least, hacking is exactly about figuring out the why and the how -- the practical result is just a pleasant side benefit.
    – jscs
    Feb 2, 2013 at 8:20

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .